main

ascend-robotsupport return routed experts & add qwen3-30b-a3b script and doc

e186ae71创建于 4月27日历史提交

文件	最后提交记录	最后更新时间
0001-apply-slime-v0.5.7-patch.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0002-Fix-qwen3-error-when-enable-dp-lm-head-16115.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0003-Fix-rotary_embedding.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0004-GLM4.7-Flash-17247.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0005-bugfix-fix-for-dsv3.2-and-dsvl2-17007.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0006-fix-Refactor-register_image_processor-to-use-kwarg-i.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0007-Support-GLM-4.7-Flash.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0008-GLM-4.7-Flash-support-graph.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0009-rotary-embedding-use-mrope-instead-native.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0010-GLM-4.7-Flash-support-prefix-cache-and-chunked-prefi.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0011-Optimize-graph-padding-for-GLM-4.7-Flash.patch	[patch] add glm4.7 sglang patch Co-authored-by: zjchenn<zjchenn@gmail.com>	3 个月前
0012-Fix-sglang-slime-qwen3_vl.patch	Fix Accuracy in Qwen3 Vl Co-authored-by: liu_zhi_xu<liuzhexu1@huawei.com>	3 个月前
0013-support-return-routed-experts.patch	support return routed experts & add qwen3-30b-a3b script and doc Co-authored-by: liyongwen<1310439159@qq.com>	2 个月前